US3071694A

US3071694A - Signal translating device

Info

Publication number: US3071694A
Application number: US402858A
Authority: US
Inventors: Theodore H Bonn; Robert D Torrey
Original assignee: Sperry Rand Corp
Current assignee: Sperry Corp
Priority date: 1954-01-08
Filing date: 1954-01-08
Publication date: 1963-01-01
Anticipated expiration: 1980-01-01

Description

1, 1963 '1'. H. BONN ETAL SIGNAL TRANSLATING DEVICE 14 Sheets-$heet 2 Filed Jan. 8, 1954 Q 0.. 4 B B A R .l. G W T. R m K m u E E c N P w v 0 G T O E L l U P RB S O 6 lllllul 6L 5 t Y 4 lllll' t 3 l I. t 2 |l|.l.|| l||||| 1 V I'll llllllllullll. t O O O O INVENTORS THEODORE H. BONN ROBERT D. TORREY ATTORNEY Jan. 1, 1963 Filed Jan. 8, 1954 T. H. BONN ET AL SIGNAL TRANSLATING DEVICE 14 Sheets-Sheet 3 Fig. 60

CK Flg. 6

1 REVERTING 1 a o A A SIGNAL POWER 0 POWER wmome i BLOCK POWER 0 "1 2 '5 1 BLOCK c 1 REVERT H (/V 0 81 RS '1 2 D BLOCK A OUTPUT SIGNAL B i Fig.70

IN VENTORS THEODORE H. BONN ROBERT D. TORREY Jan. 1, 1963 T. H. BONN ET AL SIGNAL TRANSLATING DEVICE Filed Jan. 8, 1954 14 Sheets-Sheet 4 3L v POWER L F o B I 2 LOCK R5 D A OUTPUT BLOOK 0 B1 REVERT o K v EH 0uTPuT Fig.

SIGNAL POWER BLOCK Fig. 90

D THYRITE OUTPUT c Fig.1O

INVENTORS THEODORE H. BONN ROBERT D. TORREY A TTORNE Y l, 1963 T. H. BONN ETAL SIGNAL TRANSLATING DEVICE 14 Sheets-$heet 5 Filed Jan. 8, 1954 R K E m w a L P t G 1 V O T F U P T U 0 M R 11 Dc II B d 1 a NE g 2 all Fig. 120

R EVERTING Fig.12

SIGNAL Fig.13o

QUTPUT INVENTORS THEODORE H. BONN ROBERT D. TORREY A TTORNE Y T. H. BONN ETAL SIGNAL TRANSLATING DEVICE Jan. '1, 1.963

14 sheets-sheet 6 Filed J 8.n. 8, 1954 OUTPUT INPUT PPQ i L OUTPUT Fig. 14

INVENTORS THEODORE H. BONN ROBERT D. TORREY Jan. 1, 1963 T. H. BONN ETAL 3,071,694

SIGNAL TRANSLATING DEVICE Filed Jan. 8, 1954 14 Sheets-Sheet s B BLOCK POWER BLOCK OUTPUT REVERTING OUTPUT t t t t t t t FROMD POWER PULSE SET m- OUTPUT REVERTING Fig. 18

POWER RESTORE .l D BLOCK T D1 /2 SETo c :OUTPUT V4 RL z R1 J T? REVERTING R D5 AND BLOCK I W l f l I I DELAY 0 INVENTORS f 7 THEODORE H. BONN ROBERT D. TORREY Fig. 180

Jan. 1, 1963 T. H. BONN :ETAL 3,071,694

SIGNAL TRANSLATING DEVICE Filed Jan. 8, 1954 14 Sheets-Sheet 9 Fig.18b

POWER BLOCK 0 SET RESTORE REVERTING o BLOCK 0 OUTPUT 2 INPUT- IC D4 F|g.19

a 0UTPUT BLOCKING CIRCUIT v POWER 0 INPUT BLOCK OUTPUT t1 *2 F t t t t INVENTORS THEODORE H. BONN 190 ROBERT D.TORRE Y ATTORNEY Jan. 1, 1963 1-. H. BONN ETAL 3,071,694

SIGNAL TRANSLATING DEVICE Filed Jan. 8, 1954 14 Sheets-$heet 1O INPUT POWER 1| 0 Fl X A" BLOCK :OUTPUT C RL :s e: 19C OUTPUT INPUT Fig. 20

IN V EN TORS THEODORE H. BONN ROBERT D. TORREY ATTZNEY Jan. 1,, 1 963 T. H. BONN ET AL 3,071,694

SIGNAL TRANSLATING DEVICE Filed Jan. 8, 1954 14 Sheets-Sheet 11 Fi zo d Fig. 20b

POWER REVERTING OUTPUT POWER REVERTING INPUT INVENTORS OUTPUT THEODORE H. BONN R RT D. TORREY Fig.2l0

ATTO EY Jan. 1, 1963 '1'. H. BONN ETAL SIGNAL TRANSLATING DEVICE l4 Sheets-Sheet 12 Filed Jan. 8, 1954 POWER REVERT SIGNAL OUTPUT Fig. 22a

OUTPUT RESTORE Fig. 230

R E W O P l I I INVENTORS RESTORE THEODORE H. BONN ROBERT D. TORREY OUTPUT TTO EY Jan. 1, 1963 T. H. BUNN ET AL 3,071,694

SIGNAL TRANSLATING DEVICE Filed Jan. 8, 1954 14 Sheets-Sheet 13 POWER PULSE BLOCK 1 1 fDg, T 2Jfi/' 2A/" OUTPUT F RL INPUT 0-4 C E r D2 0 D4 -.-.I

k1 L1 F g- 2 4 Q 7: E

c A BLOCK +E POWER l E BLOCK +E l I BLOCK I SlGNAL r I T 4 CAPACITOR INVENTORS THEODORE H. BONN ROBERT D. TORREY o t OUTPUT a W T TTOR Y Fig.24"

Jan. 1, 1963 T. H. BONN ET AL SIGNAL TRANSLATING DEVICE Filed Jan. 8, 1954 14 Sheets-$heet l4 POWER 1 l a REVERTING [D1 D5 1 --fi w---v R3 OUTPUT POWER REVERTING SIGNAL Fig. 24d

CAPACITOR F OUTPUT INVENTORS THEODORE H. BONN ROBERT D. TORREY United States PatentO 3,071,694 SIGNAL TRANSLATING DEVICE Theodore H. Bonn and Robert D. Torrey, Philadelphia,

This invention relates to signal controlled translating devices and more particularly to signal controlled translating devices utilizing magnetic phenomena for producing the desired translation and to logical circuits for the use thereof.

In the operation of electronic computing equipment and the like, operating at extremely high speeds, amplifiers are used to perform various logical functions. Electron tubes employing the valving action of a charged grid on an electron stream have been satisfactorily used for amplification purposes. However, a disadvantage in the use of electron tubes is the rather limited like thereof due to failure of such tubes and the fragile construction thereof. It has been proposed to replace the electron tube ampifier elements with solid-state elements that will not be subject to the limited life defects of electron tubes.

Ferromagnetic materials may be employed in such amplifying apparatus. They exhibit a hysteresis loop and display a high impedance when operating over the portion of the loop from minus residual flux density to plus residual flux density and show a low impedance when traveling from plus residual flux density towards plus saturation flux density. Use can be made of these effects for signal translating and amplifying purposes. A way of using this effect is to produce the desired output when and while the core occupies the low impedance portion of its hysteresis loop. The present invention covers devices using this efiect and may be conveniently referred to as series signal translating or amplifying devices.

With working frequencies in the megacycle region, many problems are encountered. They may deal, for example, with the speed with which a material can traverse its hysteresis loop, with the transmission of unwanted energy to the input circuits and with the generation of spurious outputs.

It is, therefore, a primary object of the invention to provide new and novel signal translating apparatus.

Another object of the invention is to provide new and novel signal translating apparatus operating in series.

An additional object of the invention is to provide new and novel signal translating apparatus which will operate successfully at high frequencies.

A further object of the invention is to provide new and novel signal translating apparatus employing magnetic phenomena.

Another object of the invention is to provide new and novel signal translating apparatus having an eflicient and fast operating signal input arrangement.

An added object of the invention is to provide new and novel signal translating apparatus having an improved output arrangement to suppress unwanted outputs.

Still another object of the invention is to provide a new and novel signal translating apparatus utilizing only one coil on a core of ferromagnetic material.

An additional object of the invention is to provide new and novel signal translating apparatus having a steady output when pulses are applied to the input.

3,071,694 Patented Jan. 1, 1963 An added object of the invention is to provide new and novel complementing apparatus employing magnetic phenomena.

Another object of the invention is to provide new and novel signal translating apparatus which will operate to interlace signals.

A further object of the invention is to provide new and novel signal translating apparatus which will operate as a multi-channel distributor.

Still another object of the invention is to provide new and novel signal translating apparatus functioning as a frequency divider.

An added object of the invention is to provide new and novel signal translating apparatus to operate like a flip-flop.

A further object of the invention is to provide new and novel signal translating apparatus which will function as a pulse counter.

An additional object of the invention is to provid new and novel complementing apparatus operating as a shifting register for use as a delay line or for stepwiseincreasing the power gain.

Other objects of the invention will in part be described and in part be obvious as the following specification is read in conjunction with the drawings in which:

FIGURE 1 is a diagram of an idealized hysteresis loop;

FIGURE 2 is a schematic showing of an embodiment of the invention;

FIGURE 2a represents the operating time cycle of the embodiment of FIGURE 2;

FIGURE 3 is a schematic showing of a modification of the invention;

FIGURE 3a represents the operating time cycle of the embodiment of FIGURE 3;

FIGURE 4 illustrates some representative output waveforms of the embodiment of FIGURE 3;

FIGURE 5 illustrates some representative power pulse waveforms for the embodiment of FIGURE 3;

FIGURE 6 is a schematic diagram of a modified input circuit;

FIGURE 6a illustrates the operating time cycle of the embodiment of FIGURE 6;

FIGURE 7 is a schematic showing of a further input circuit;

FIGURE 7a represents the operating time cycle of the embodiment of FIGURE 7;

FIGURE 8 is a schematic showing of another input circuit;

FIGURE 8a represents the operating time cycle of the embodiment of FIGURE 8;

FIGURE 9 is a schematic showing of a modified output circuit;

FIGURE 9a represents the operating time cycle of the embodiment of FIGURE 9;

FIGURE 10 is a schematic showing of a further output circuit;

FIGURE 11 is a schematic showing of another output circuit;

FIGURE 11a represents the operating time cycle of th( embodiment of FIGURE 11;

FIGURE 12 is a schematic showing of another em bodiment of the invention;

' FIGURE 12a represents the operating time cycle or the embodiment of FIGURE 12;

"FIGURE 13 is a schematic showing of a modificatiol of the embodiment of the invention of FIGURE 12;

' different properties.

FIGURE. 1341 represents the operating me cycle 9 the embodiment of FIGURE 13;

FIGURE 14 is a schematic showing ofan additional output circuit;

FIGURE 15 is'a schematic showing of another output circuit;

FIGURE 16 is a block diagram of a circuit incorporating the invention;

FIGURE 1611 represents the operating time cycle for the circuit of FIGURE 16;

FIGURE 16b is a schematic showing of the circuit of FIGURE 16;

FIGURE 17 is a schematic showing of a frequency divider;

FIGURE 17a represents the operating time cycle for the circuit of FIGURE 17;

' FIGURE 18 is a block diagram of an arrangement utilizing the invention for flip-flop effects;

FIGURE 18a is a schematic diagram of the arrangement of FIGURE 18;

FIGURE 18b represent-s the operating time cycle for the circuit of FIGURE 18;

FIGURE 19 shows in schematic form a scale-of-two counter;

FIGURE 19a represents the operating time cycle for the circuit of FIGURE 19;

FIGURE 19b shows in schematic form another scaleof-two counter;

FIGURE 19c represents the operating time cycle for the circuit of FIGURE 19b;

FIGURE 20 is a schematic showing of a shifting register delay line;

FIGURE 20a represents one operating time cycle of the circuit of FIGURE 20;

FIGUREZOb represents another operating time cycle of the circuit of FIGURE 20;

FIGURE 21 is a schematic showing of an additional embodiment of the invention;

FIGURE 21a represents the operating time cycle for the circuit of FIGURE 21;

FIGURE 22 is a schematic showing of a modification of the embodiment of FIGURE 2 1;

FIGURE 22a represents the operating time cycle for the circuit of FIGURE 22;

FIGURE 23 is a schematic showing of a circuit utilizing the embodiment of FIGURE 21 for flip-flop effects;

FIGURE 234 represents the operating time cycle for the circuit of FIGURE 23;

FIGURE 24 gives a block diagram of two signal translating devices connected in cascade with a common power pulse input and a common blocking pulse input, and with a capacitive delay interposed between them;

FIGURE 24a is a schematic showing of one embodiment of the invention based upon the principle illustrated in FIGURE 24.

FIGURE 24b represents the operating time, cycle for the circuit of FIGURE 24a.

FIGURE 240 is a schematic showing of another embodiment of the invention basedupon the principle illustrated in FIGURE 24.

FIGURE 240. represents the operating time cycle for the circuit of FIGURE 240.

FIGURE 1 illustrates an idealized hysteresis. loop of a material which may be used as the core member for the solid-state signal translating devices to be described. 'B signifies residual flux density, and B designatessaturation flux density. The core material may be made of a variety of materials amongst which arethe various types of ferrites and the various kinds of. magnetic tapes, including Orthonik and 4-779 MolyPermalloy. These materials may have different heat treatment-sv to give them In addition to the wide variety of materials applicable, the cores of the signal translating devices may be constructed in a number of different geometries. involving both. closedand open paths. For. ex-

cores are possible.

It is to be understood that the invention is not limited to any specific geometries of the cores nor to any specific geometries of the cores nor to any specific materials therefor, and that the examples given are illustrative only. The only requisite is that the material possesses a hysteresis loop preferably approaching the idealized hysteresis loop as shown in FIGURE 1.

As to the different kind of electric pulses to be applied, the following terminology has. been used: There are clock pulses and signal pulses. The signal pulses carry information and are, therefore, selectively applied.

such pulses are present or not. The clock pulses, in contrast thereto, are automatically applied at fixed intervals and do not carry any information. They may be subdivided into power pulses and blocking pulses. The power pulses usually supply the power for the operation of the translating device or, at least, open a gate to permit another source to operate the device. The blocking pulses block the interference of the power pulse with the signal input circuit and/or of the signal input circuit with. the power circuit.

FIGURE 2 is a simplified schematic diagram of the elements of a magnetic signal translating device. Part C is a core of ferromagnetic material. Winding 1 is the input or signal winding, while winding 2 is the output or power winding. Reference is made to the idealized hysteresis loop of FIGURE 1 to assist in describing the operation of this signal translating device. Waveform B, FIGURE 2a, is applied to'the power winding through terimnal B, its positive half, as, for example, during the period t to t representing the power'pulse, and its negative half, as, for example, during the period t to t representing a blocking pulse which prevents any current in winding 2 from flowing through diode D during the sig nal periods, which are the periods t to t or 1 to The signal is selectively applied during the signal periods at terminal A. During the power pulse periods, which are the periods to or i to t or t to t a blocking pulse is applied to the signal winding at terminal B thus blocking diode H and preventing any current flow through the signal winding. As a result, diode D is conducting only during the power pulse periods, and diode H is conducting; I

only during the signal periods.

During the power pulse periods, the positive half of waveform B, the power pulse, establishes flux in the core in the direction of the solid arrow shown at the bottom end of the core, and the core, assumed to be at plus B at time travels from plus B to plus B which represents. the low impedance region. An output signal is the result of this operation, and the core returns to plus B If. a.

signal is applied during one of the signal periods, flux is established in the core in the direction of the dotted arrow shown at the bottom end ofthe core, and the core travels from plus B to minus B which represents the high impedance region, in the counter-clockwise direction. The next following power pulse will set the core back to plus B but does not produce an output pulse. It can easily be seenthat this device always yields an output pulse in response to the power pulse, except when an input has occurred on terminal A during the preceding signal period. The device, therefore, operates as a complementer because the intended output signal is produced in response to the non-application of an input signal.

Attention should be given to the fact that the device illustrated in FIGURE 2' as well as all the other devices described hereinafter represent so-called series translating devices. This means that the load circuit or circuits are arranged in a series relationship to a winding associated with the core in regard to the power pulse source. As a result, the intended output signals are pro duced, as a rule, through the excursionof the magnetic.

It depends upon the information to be transmitted whether core from plus residual flux density towards plus saturation flux density, as illustrated in the case of FIGURE 2.

t also should be realized that not only the device of FIGURE 2, but every series translating device, as described in this specification, operates as a complementer in reference to what may be called a reverting pulse. Such a freverting pulse causes the magnetic core to be flipped from plus B to minus E or, more generally speaking, from the position on the loop, which the core has acquired as a result of the power pulse, to the very opposite position. There is no desired output pulse, whenever a reverting pulse has been applied, because the next following power pulse finds the core in the minus B position and drives it through the high impedance region of its hysteresis loop, the result of which is to leave the core at plus B The second following power pulse will then cause the core to travel from plus B over the low impedance portion of its hysteresis loop towards plus B provided that there was no interfering reverting pulse, and an output pulse will result in response to the non-application of the reverting pulse.

If it is desired to use the device as an amplifier, and not as a complementer so as to produce the intended output signal in response to the application of an input signal, then this input signal cannot be employed as a reverting pulse. The reverting pulse must be applied by a source which is not the signal source, and the signal itself must be used to inhibit the passage of the reverting pulse through the signal winding. It is, therefore, only through a double negation, that a series translating device can be converted from a complementer into an amplifier,

It should be understood, however, that the term amplifier, as used here and hereinafter in this specification, is not confined to cases of actual amplification, but extended to cover all devices which furnish the intended ou put signal in response to the application of an input signal, regardless of the fact that the power, current or voltage ratio may be greater than, equal to or less than unity.

FIGURE 3 shows a signal translating device which, in accordance with the foregoing explanations, illustrates the conversion of the complementer of FIGURE 2 into an amplifier. The Winding 2 in FIGURE 3 corresponds exactly to the winding 2 in FIGURE 2. The conversion from a complementer into an amplifier expresses itself in the differences between the windings l of the two drawings and the electric impulses applied to them. The electric impulses applied at terminal B in FIGURE 2 operate as blocking pulses, as explained hereinabove. Their only purpose is to prevent any current flow through the signal winding during the power pulse periods. In contrast thereto, the electric impulses applied to terminal B in FIGURE 3 operate both as blocking pulses and reverting pulses. During the power pulse period, they fulfill the same function as the blocking pulses applied at terminal B in FIGURE 2. During the signal period, they flip the core back to minus B as discussed hereinabove. Since these pulses are applied in form of clock pulses, the core is constantly flipped back to minus B and, consequently, there will be no desired output pulse as long as gate G permits the passage of these reverting pulses. The only way, therefore, to obtain the desired output pulse is to inhibit the passage of these reverting pulses through the gate G, and this is accomplished through the application of the signal pulse which is employed as an inhibiting impulse to close the gate. As a result, the desired output pulse emanates in response to the application of the signal pulse. The operating time cycle for FIGURE 3 is shown in FIGURE 3a.

As long as no signal pulse is applied, the material operates in the high impedance portion of the hysteresis loop, and there is very little output. A typical form of this kind of output is shown in waveform X, FIGURE 4. This output occurs during the power pulse periods and is cal-led a sneak output or sneak pulse.

is the low impedance portion of the loop, and the core will present a low series impedance. An output form such as curve Y, FIGURE 4, will appear across R Then, at the end of the power pulse period, the material returns to plus B After that, it may be sent to minus B by another reverting pulse at B or the passage of the reverting pulse may be inhibited, as desired, and another output pulse obtained.

The conversion of the complementer of FIGURE 2 into an amplifier may also be accomplished in the following manner: The polarity of the output is to be inverted so as to obtain negative output pulses at all times, except when there is a signal; the output will then be zero. If then a positive direct current shift is effected by an amount equal to the amplitude of the negative output pulses, the complementer has been transformed through double negation into the equivalent of an amplifier.

The power delivered to the load may be many times larger than the power required of the information pulse A net power gain is, therefore, obtainable in the signaltranslating device. Many factors influence the amount of power obtained. One of the most important factors, however, has to do with the extent to which the unwanted pulse, known as the sneak pulse and shown at X in FIG- URE 4, may be tolerated in any practical situation. Another important factor is represented by the ratio ,of the slope on the steep portion of the hysteresis loop between plus B and minus B to the slope of the fiat portion of the hysteresis loop between plus B and plus B A material with a rectangular hysteresis loop is desirable for this signal translating device, although by no means completely necessary.

In FIGURE 3a, the Waveforms applied to terminals B and B are mirror images with respect to the time axis. This condition need not be rigidly met, however; their amplitudes may difler, depending on the winding constants, and their durations may differ, depending on the degree of blocking overlap considered necessary or desirable. During the power pulse periods, the voltage applied to B should be equal to or greater than the amplitude of the pulse in the signal Winding resulting from hys-' teresis loop traverse during the power pulse, and during the signal periods, the voltage applied to B should be sufficient in amplitude to block the flow of current in the output winding.

Actually, t r, and I and so on mark the boundaries of the periods allotted to the power and signal pulses and indicate by no means the length of these pulses. The

period between 1 and t;;, a signal pulse period, can be greater or less than the period 1 to t a power pulse period. Because of the selectibility of the lengths of the respective pulse periods, this signal controlled translating device may also serve as a memory or as a delay device.

In the foregoing descriptions, rectangular waves have been shown for the various power, signal, reverting and blocking voltages. This is not required. The only requirements on the variouswaveforrns are: (1) The blocking voltage must be at least as large as the pulse which it is blocking. (2) The pulses applied to the windings which cause the core to traverse portions of its hysteresis loop must be such that the integral of these voltages with respect to time is selected to be large or small in accordance with the size of the output signal which is desired.

In view of the fact that the power pulse is derived from a source whose waveform may be accurately fixed, output pulses from this amplifier are of standard waveforms as determined by the power pulse source. Therefore, this amplifier serves also as a pulse former.

FIGURES shows some typical shapes of power pulses which might be used. FIGURE A illustrates a half sine wave; FIGURE 51) pictures a triangular wave; FIGURE 50 exemplifies a pulse with an initial high amplitude peak and a steep fall to an even-leveled portion which may be regarded as a preferred wave form; and FIGURE 5a? shows a flat-top pulse with unequal rise time and fall time.

The description has been directed to the core material being reverted from plus B to minus B Actually, it is not necessary to operate over the full range of B The material could operate, for example, from plus B down to plus A B Operation over part of the hysteresis loop will bring decreases in power without decreasing the volume of core material involved.

It is also possible to operate the core material over a range greater than minus B to plus B In such case, the integral of the voltage with respect to time of the power pulse must be greater'than the value required to cause the core to travel from minus B to plus B and will cause the core to travel beyond plus B toward plus B This operation would otherwise require two power pulse periods, one in which the core is operating from minus B to plus B and one to travel from plus B to plus B In this case, an output pulse will always appear, and the duration of the output pulse will depend upon whether the core had been at minus B or plus B Several input circuits will now be shown to operate this amplifier with both constant current and constant voltage sources. A constant current source is theoretically a source of infinite impedance. A constant voltage source is theoretically a source of zero impedance. These definitions are idealized and are merely used to obtain a simplification in the analyses of circuits. From a practical point of view, the constant current source is a source whose impedance is comparatively high with respect to the load, and a constant voltage source is a source whose impedance is comparatively low with respect to the load.

FIGURE 6 shows a method of using a low impedance voltage source for the pulses on the input winding. A voltage as shown in the accompanying waveform diagram of FIGURE 6a is applied to terminal B from a low impedance voltage source. In this case, t to t is the signal period. If an output is desired, a signal pulse, as illus trated', is applied to terminal A. This prevents source B from sending the core from plus B to minus B A power pulse is shown in the bottom waveform of FIGURE 6a. Period t to t is the power pulse period. It goes without saying that the diodes, the direction of winding of coil 1 and the polarity of the applied voltages may be reversed in a similar way as is discussed, hereinafter, in connection with FIGURES 7, 7a, 8 and 8a.

FIGURES 7 and 8 illustrate two input circuits using a high impedance voltage source for the input. C is a core of ferromagnetic material with which coils 1 and 2 of an amplifier are associated. The waveform applied to terminal B is the familiar alternate power and blocking voltage that has been previously described. In the example of FIGURES 7 and 7a, terminal B of the sig nal winding circuit is supplied during the period t to Z with a negative block. During the period t to t terminal B goes positive. R is a high resistance sutiicient tomake source B appear like a constant current source in respect to the amplifier. The signal input terminal is maintained at a positive potential. It a signal is desired, afnegative pulse is applied to terminal A to decrease the potential at terminal A to O. This selectively prevents the application of the current from source B to coil l, and a large output will result in the succeeding power pulse cycle. This particular arrangement of polarities provides that the amplifier produces a positive output pulse when a negative pulse is applied to its input. The input signal polarity can be changed by reversing the diodes H and H reversing the direction of winding of coil, reversing the polarities of the pulses applied at B V and shifting the signal voltage level, or the output pulses can be changed by reversing diode D and the polarities of the pulses applied at B. The arrangement for revers ing the input signal polarity is shown in FIGURE 8 together with FIGURE 8a. This arrangement provides a positive output when a positive pulse is applied on the input. The power period is t to t and t to I is the signal period, as in FIGURE 7a. However, the direction of winding of coil I in FIGURE 8 is opposite to the direction of Winding of coil 1 in FIGURE 7, and the polarity of the voltages applied to terminals B and A is reversed. If an output is desired, terminal A has a positive pulse applied thereto to raise the potential at terminal A to 0 during the period 1 to t;,, as shown in FIG- URE 8a. Now the amplifier has both positive input and positive output.

The various equalities of voltages which have been specified apply only to the cases where there is a unity turns ratio between power and input windings; If the turns on these windings are not equal, as may be the case, the turns ratio will multiply one side of the equality. Also, in this amplifier, the description is based upon a particular set of pulse polarities. It should be noted that either the signal polarities or the power pulse polarities or both could be reversed by reversing the direction of the winding.

One of the matters of concern in connection with the above-described amplifiers is the removal of the sneak output or sneak pulse. FIGURE 9 shows a circuit for suppressing the sneak pulse. Winding 2 is the output coil of a magnetic amplifier. A positive power pulse is applied to terminal B. At the same time, a negative pulse is applied to terminal L, as shown in FIGURE 9a. This negative pulse is of such magnitude that when no output is expected from the amplifier, point B is at zero potential. In other words, the voltage at point L is made equal to the sneak current times R. Then, when an output is desired, point E assumes a positive potential of the waveform applied to terminal B.

FIGURE 10 shows a discrimination method using thyrite in place of the diode K shown in FIGURE 9. The resistor with the X through it is a thyrite resistor. While its resistance is very high for the low voltages of the sneak pulse, it is very low for voltages of the order of the output pulse.

FIGURES 11 and 11a illustrate another form of a sneak pulse suppressor circuit. Coil 2 is the output coil of a magnetic amplifier. Power and block voltages, as shown, are applied at terminal B. The resistor R carries a current equal to the sneak pulse current which flows from ground through diode M and resistor R to the source minus E. When the output from the ampliher is merely the sneak output, the sneak current is applied through diode D in series with R, and the output voltage remains approximately zero. However, when the desired output pulse is present, diode M disconnects, and the output voltage jumps up to the voltage of the power pulse. to the sneak suppressor circuit, i.e., R in conjunction with E, an amount of power equal to the output voltage times the sneak pulse current. Since this is approximately equal to the power required to set the core, this sneak pulse suppressor circuit subtracts from the power output an amount equal to the power of the input. It should be noted that, if the amplifier is to feed any gate circuits, the resistor R can carry all of the gate currents which must be supplied by the amplifier.

It is possible to construct an amplifier utilizing only one coil associated with a core of ferromagnetic material. The advantage of this type of amplifier is that it would be easier to construct than an amplifier with two coils. The limitation is that this type of amplifier. has current again, but no voltage gain. In this respect, it is similar to the vacuum tube cathode follower and could be called a magnetic cathode follower.

FIGURE 12 is a schematic showing of a single coil In this condition, the power pulse supplies magnetic amplifier. A single amplifier coil is associated with a core of ferromagnetic material C. The waveform shown in the timing diagram of FIGURE 12a is applied to terminal B while terminal B is held at Zero potential. During the period t; to t terminal B is positive, and current flows from terminal B through the amplifier winding, diode D and resistor R to ground. Referring to the hysteresis loop of FIGURE 1, the core is moved from minus B to plus B During the period t to terminal B is negative with respect to ground, and current flows from terminal B through resistor R diode D and the amplifier winding to terminal B. Note that, in general, RS will be different from R and generally will be larger. Therefore, the positive and negative halves of the waveform applied to terminal B are not necessarily equal in magnitude.

When operated in this condition, no output results from the amplifier. If an output is selectively desired, the waveform shown in FIGURE 12a is applied to terminal A. This is a negative voltage equal to or greater than the negative voltage applied to terminal B during the period to i This blocks diode D and prevents the core from being operated from plus B to minus B The following positive pulse applied to terminal B will then find the core material in a low impedance state, and a large output will result. Since R is in general larger than R a current gain is obtained from this amplifier. The essential feature of the D R circuit and the input terminal A is that this circuit is a means for inhibiting the negative reset pulse applied to the coil. Other methods to do this will be described hereinafter.

FIGURE 13 shows a modification of the single coil magnetic amplifier of FIGURE 12. Here the operation of the core is the same as described for FIGURE 12. The core material operates between minus B and plus B if no power output is desired, or between plus B to plus B if a power output is desired. R is a high impedance, relative to R and the waveforms as shown in FIGURE 13a are applied to'terminals B and B The voltage applied to terminal B is sufficient to allow a current equal to the magnetizing current required by core C to flow through the coil. Between t and t terminal B goes negative by an amount equal to the voltage developed in the coil when the core flips. Point E is then at zero potential during the period'when core C flips from plus B to minus B Therefore, all that is required to inhibit this flipping is an input voltage at terminal A which is equal to or greater than the negative voltage applied to B. This will block diode D The current which fiows into the signal source at A through diode D is equal to the current through R which, in turn, is equal to the flipping current. Therefore, it can be seen that, with this type of input circuit, the amount of control power is equal to the power required to set the core from plus B to minus B and that this is the minimum that could possibly be required. Of course, the signal is applied to A onlywhen an output is desired. It should be noted that the reason for applying the negative voltage to B during the period t 't0 i is to keep point E from rising aboveground potential during that period. This blocks diode D and prevents the output circuit from interfering with the input circuit.

In the magnetic amplifiers previously described, it is sometimes desirable to obtain a steady output when pulses are applied to the input. This can be done by utilizing a rectifier or suitable filter or integrating circuits at the output.

FIGURE 14 exemplifies a circuit for obtaining a steady output. The output coil of a magnetic amplifier, as illustrated in the arrangement of FIGURE 3, is shown at 2. The output pulses charge up the capacity F which discharges slightly during the output pulse periods, and a steady output is obtained. If it is desired toreduce the steady output to zero, the input pulses to the amplifier are removed. It may take several pulse periods for the charge to leak off capacity F. If a faster decay is desired, a clamp pulse could be applied to the output to reduce the output rapidly to zero.

Another circuit to obtain a steady output is shown in FIGURE 15 which utilizes an ordinary delay line as a pulse stretcher. The circuit shown with diodes D D and D feeding various points on the delay line is a standard type of a pulse stretcher. It utilizes the propagation time from the various points on the delay line to the output to increase the length of the pulse. In an amplifier which operates on a 50-50 duty cycle, i.e., one in which the signal period and power period are of equal length, only two-diodes would be necessary. The delay time between the two diodes would be equal to the length of either the signal or power periods. By the time the main output pulse injected at the output end of the delay line would have died down, the pulse from the diode down the transmission line, which is delayed by a time equal to the power pulse period, will have arrived at the output, and it will last for a period equal to the signal period. Therefore, a steady output will be obtained when pulses are applied to the delay line. This type of circuit has the advantage of a much more rapid fall time than the capacitor circuit shown in FIGURE 14 and, in general, no reset pulse will be required with this type of a circuit.

Of course, where the pulse periods are not the same, for example, where the signal period is longer than the pulse period, the same type-of circuit could be used. However, it would be necessary to employ more than two diodes. In the circuit of FIGURE 15, three diodes are shown. This could be used for any signal period length up to the point where the signal period is twice the power period. It is evident how this principle can be extended to include signal periods of any duration.

FIGURE 16 shows a method of interlacing signals in such a manner that the rate of transmission of information over one single line can be doubled. This means that, without any change in the speed of propagation, the time required for the transmission of a given quantity of'information is cut into one half. To accomplish this effect, a pulse envelope system may be used, as illustrated in the timing diagram of FIGURE 16a. The timing pulses (TP) and power pulses (PP) are shown in relation to the input and output pulses. The elements A and A are magnetic amplifiers as previously described. The timing pulses TF and TF alternately gate the applied direct current into amplifier A and A during the time intervals preceding the respective power pulse periods to produce the reverting effect described hereinabove. As a result, the power pulses PP and PP which are alternately applied to either amplifier do'not drive the respective cores into the low impedance region and do not effect, therefore, any output. The input signal inhibits the passage of the direct current through the common inhibitory gate, and A and A alternately produce outputs during such power pulse periods which are not preceded by the application of a reverting pulse. These outputs are then bufied into a common output line.

It is to be noted that an interlace of more than two circuits could be used. Three, four or any desired number of gates, amplifiers and buffers could be connected in the same manner as the two circuits shown in FIG- URE 16 are connected. A schematic diagram for a two circuit interlace is shown in FIGURE 16b.

This circuit may also be used for the purpose of rearranging the pulses in a group. The pulses are fed into the appropriate amplifiers, as determined by the timing pulses, and are fed out at the time determined by the power pulses. The output pulses could, therefore, have a different array than the input pulses. By the proper use of multiple outputs, an interlace device such as this can be used, furthermore, to take signals arriving on a single input line and distribute them at the proper times,

asdetermined by the timing and power pulses, to several output lines.

A complementer operating as a frequency divider is shown schematically in FIGURE 17. The timing diagram of FIGURE: 17a assists in explaining. the operation. No input signal, but only a blocking pulse is applied to terminal B of the input winding. Assuming the core to be at plus B when a power pulse is fed to terminal B at time t an output pulse. is produced which, beside being fed to the output terminal, is fed to the input of a delay unit D which may consist of an amplifier, a delay line or some low pass circuit; The resulting delayed pulse is applied as a reverting pulse to winding 1 at. time t to drive the core. to minus B When the next power pulse isapplied at time t;,-, there will be no intended output from: this complementer as the core travels from minus B to plus B and no pulse will be applied to winding 1 from the delay unit at time 12;. This sequence then repeats itself beginning with the application of the next power pulse at time 1 Therefore, this complementer acts as a frequency divider.

FIGURE 18 shows in block form how to usean am plifier for flip-flop eifects. The corresponding schematic and timing diagrams are found in FIGURES 18a and 18b, respectively. A set signal prohibits the reverting pulse from driving the core to minus B and leads, therefore, to the production of an output signal which, besides being applied to the output line, is fed toadelay unit which will delay the passage of the pulse therethrough for a length of time equal to the power pulse period. The delay unit D'may be a delay line, another amplifier or a low pass circuit or a capacitance or inductance. The output of the delay unit D is then butfe'd onto the input line of the amplifier, after passing through the inhibitory gate G, to permit production of another output pulse during the next power pulse period. In this manner the circuit of FIGURE 18 will continue to produce output pulses. If the feedback loop'tothe input is broken for the time duration of the input period, the amplifier will no longer produce output pulses, and the flip-flop will be restored. Thisis accomplished by the inhibitory gate G with a restore signal being applied to inhibit passage of the delayed output pulse. No output results then until a set signal is again applied tothe input terminal.

In the schematic of FIGURE 180, the positive half of the blocking voltage acts as a reverting pulse and keeps the amplifier in the high" impedance state by a current ilow through winding .1 and resistor R to ground. If, however, a setting signal is applied, diode D raises the voltage across R to block this action and to leave the core C. inits low impedance state. This results in an output on the next power pulse. The output is then delayed for the length or" the power pulse through the delay line, and, during the next signal period, this delayed output pulse is applied to winding 1 through diode D causing the same effect as the set input. Thus, there will be a continuation of output pulses. To restore the flip-flop, a negative signal is applied through diode D opening the path of the circulating output pulse and allowing the amplifier to reset itself to the high impedance state again, and, thus, no more output pulses are produced.

To obtain a continuous output rather than a pulsed output, the output from the delay device D may also be buffed into the output line. In this manner, the delayed output pulse is not only fed back to keep-the flipflop set, but is also applied to give an output pulse during the time of the block pulse, thus maintaining a steady output as long as the flip-flop is set. A restore signal will then inhibit any further output signal, and the flip-flop will be rapidly restored.

A single winding scale-of-two counter is shown in FIGURE 19 with the corresponding timing diagram in FIGURE 19a. At time 23, the input goes positive (power 12 pulse), and diode D iscut off due to the fact that there is a' higher potential: at: point-X= than at point Y. Assuming the core to be at plus- B fiux density at time r an output results and capacitor F is charged. At time t the input goes negative and disconnects diodes D and D (blocking pulse). At the same time, a blocking circuit prohibits any flow of current to the load. As a result,,- capacitor F discharges through the amplifier coil, diode D and resistor R to ground. This reverse current flow through the amplifier coil causes-the amplifier core to travel to minus B Thus, when the input goes positive again at time t the core is caused to iiip to plus B but, since only a sneak outputresults, capacitor F receives only a small charge. Attime 14, when the input signal againgoes negative, there is no change in conditions. The cycle of operation: repeats again, beginning at time t It thus requires two input pulses for each output pulse.

Inplaceof the capacitor in FIGURE 19, the use of an artificial transmission or delay line or pulse forming network may well be'used; The use of such devices permits av control of the shape of the output pulses an'd'of the core flipping pulsesas'well as a more efi'icient'discharge of the energy stored through the winding.

Another fornr of an amplifier operating as a counter is shown'in FIGURE 19b with the corresponding timing diagram of FIGURE 1%. The input wave consists, in accordance with the illustration, of a positive power pulse. which may be either a clock pulse oraselectively applied information signal, and a negative blocking pulse which cuts oif diode D1. Assuming th'e core to beat plus B at time t when the power pulse is applied, the core travels to plus B and then return's'to plusB This results in the desired output signal, as shown in FIGURE 190. At time t the diode D is disconnected, and the energy of the magnetic field causes the coil to charge the condenser F. This condenser then discharges back through the coil, setting the core to minus B ibefore the time i (reverting pulse); The next following power pulse, which sets the core to plus B is not sufficient to produce aregular output signal, but results only in a sneak pulse as shown. At time t the power pulse drives the core again to plus B and the cycle of operation repeats itself. Thus, there is always one output signal in response to two input signals, as in the case of the device illustrated in FIGURE 19. The waveform marked X in FIGURE indicates the voltage at point X.

Magnetic translating devices may be combined in the manner shown in FIGURE 20 to obtain a shifting register delay line; A plurality of magnetic complementers are connectedin chain fashion, and the information is selectively applied to the input winding of the first complementer of'the chain. The output winding of each complementeris' connected through a diode to the input windingv of the next succeeding complementer, and the output of the last complementer in the chain may be used as-the final output. It is to beunderstood, however, that outputs may be taken from any complementer in the chain. The chain may be as longas desired. The output from theregister'may also be joined to the input through appropriate'gates, as is well known in the art, to form a circulating register. a

Two examples are given for the operation of such a shifting register. In the example of FIGURE 20a combined with FIGURE 20,

lines

1 and 3 alternately pro-' vide the power pulses for alternating:complementers,

lines

2 and 4 alternately provide both the larger blocking pulses and the smaller sneak suppressor pulses in such a way that a blocking pulse in line 2 is accompanied by a sneak suppressor pulse in line 4' andvice versa, and line 5 is kept at a zero potential.

If no input signals are applied at the input to respective complementers, the complementers are maintained at plus B during the input period. During the output period, when a power pulse is applied to any comple-

Claims

1. IN COMBINATION, A CORE OF FERROMAGNETIC MATERIAL CAPABLE ASSUMING BISTABLE STATES OF MAGNETIC REMANENCE, A POWER WINDING AND A SIGNAL WINDING ASSOCIATED WITH SAID CORE, A SOURCE OF CLOCK PULSES AND A LOAD CIRCUIT CONNECTED TO SAID POWER WINDING, SAID LOAD CIRCUIT BEING IN A SERIES RELATIONSHIP TO ONE OF SAID WINDINGS ASSOCIATED WITH SAID CORE IN REGARD TO SAID SOURCE OF CLOCK